Digital BW, The Print

Ernst Dinkla wrote:

> I don't have samples like 4-6 pixels per grain but it was more to
> emphasize the issue.  I was working on 4000 PPI scan of a 35 mm Tri-X 
> film with much grain and while working on it there were things 
> happening that could only be explained as the grain interfering. I 
> didn't find it contrasty enough in a sense and either I could sharpen 
> the scan or increase the contrast. In a way the sharpening was more in 
> line with Tri-X analogue than the contrast increase. But the 
> sharpening resulted in a lighter image in total. That led to writing 
> this message. In practice it will be less of a problem than I 
> described it but I wonder how many of us are aware that there's some 
> conflict in editing grainy scans.

Yes. I'm aware, and have seen the same thing. I just accepted it as a 
"normal" problem to have to work around with Photoshop. It is definitely 
interesting to watch the histogram when you sharpen. The histogram says 
it got more contrasty, but the whole image actually seems perhaps less 
contrasty, and it does change the overall tonal balance a bit. I've had 
some images get lighter as you have. I suspect that it depends to some 
extent on the image itself and how much detail (as opposed to fairly 
featureless areas) it contains. Even then, the featureless skies may 
lighten when the entire image is sharpened, as you report.

>
> >
> There's no real problem at my end. A bit more proofing and it will be 
> alright. By trying to get the best scan it backfires more or less when 
> that scan is made of a grainy film. Which essentialy is noise but of a 
> kind that we are used to and learned to appreciate. Taking the grain 
> out one way or another (lousy scan for example) makes more garbage. 
> The print is grainy at less than A4 size so it must have been Tri-X 
> with Rodinal 1:25 or something like that, no record of it.
>
> Ernst

Since my Tri-X scans are mostly my own work, I "cheat" a little. I 
develop in XTOL 1:3 which gives me grain that is fairly small and fairly 
sharp. Then I expose/develop to give me as little density as will give 
me a good scan. In Zone System terms, N-2 more or less. The resulting 
grain in the highlights is quite small (for Tri-X) and easier to work 
with. I can easily do 10x enlargements without visible grain (from a 4x5 
negative). This helps.

Even though my monitor closely matches my prints, I still find that the 
only way to nail a result is to make proof prints. This is as much due 
to the problems you've pointed out as to the problems of scale, where a 
large print looks as if it is tonally different than a small print. In 
the end, I've still got to make full size proofs to be sure that the 
print does what I want.

But if these are my worst problems, I'm doing OK!
-- 
Bruce Watson